Ben Horowitz 谈全球科技、权力与影响力之争

When AI becomes the interface for everything, every tool we have and every technology we use is going to be built on these models. The models are not objective. They have opinions. They have strong opinions on things like history, culture, and values. And those are very different depending on where you are.

Deterrence is no longer just the size of a military. It's the pace of innovation in reacting to cheap, software-built technology that adversaries may be using, whether they're a country, a criminal, or a terrorist group. And those technologies today are not being built by governments. They're being built by the private sector.

Products and technology travel faster than companies can. So what happens is, for startup companies, usually in the olden days, you didn't care about going international until you had a few hundred million dollars in revenue. Now these companies have to go international sooner and sooner.

The number-one question we often get when we're visiting different countries is, “How do you recreate Silicon Valley locally?”

I don't think that's true, by the way. I think that's happy talk.

We are here to talk international today. For a firm that historically used to be centered in Menlo Park—and we didn't even want to open up a San Francisco office—this is a big change for us. So, Ben, why don't you take us through this announcement around our global ambition and how all the pieces between global affairs, international go-to-market, LP relations, and the global partnership level fit together?

It really starts with America, interestingly. It's been our strong belief that if you drop a person into any country in the world and they don't have the culture, money, or capital resources yet, and they just have an idea of how to make the world a better place, their best shot is in America.

That's a very special thing. A big mission of our firm is trying to preserve that special thing that is America. That means America has to be a leader. It has to be a leader economically, culturally, and militarily. To do that, we have to be a leader technologically. We have to be a tech leader. That's really our role in the ecosystem and in the world: to help America be a leader.

What we realized as we've gotten further along with this mission is that we need to bring our partners and allies into it. It started with the work that you're doing, Jen, with people from around the world wanting to participate in the U.S. technological opportunity. That meant we started getting investors from around the world.

As we went out to see those investors and said, “How can we be a better partner?” we found some interesting things. Every country wants not only to economically benefit from the technology, but also to directly benefit from it. Can you bring it to us? Can we get a chance to use it?

At a higher level, at the government-policy level, people wanted to know, “How can we be a great country for adopting technology, for innovating like you guys are, and for helping push our country into the future?” As we looked at how to step up to that opportunity and be a good partner, we realized that with our country partners, we needed relationships at the government level, the investor level, and the adoption level. That's what these 3 components become.

We've got global affairs, which builds our relationships, led by Anne, and helps us have relationships in the right way with governments ranging from Saudi Arabia to Japan to Mexico. We've had great conversations with all of them.

Then we have people who invest. Sometimes those are the governments; sometimes they're various companies in the government; sometimes they're companies that actually handle government money in the form of pensions. Finally, both the government and the businesses want to adopt the technology, and that's where Raghu's effort comes in, where we have the go-to-market component.

Anne, how much of your decision to join a16z was motivated by this mission, and also by the sense of duty in helping other countries think about the adoption of AI and not wanting to get left behind in AI, like Ben was alluding to?

Very much so. What I saw happen during my years in the intelligence community and then in the White House was technology moving from being a tool of diplomacy and national power to being the arena of it, where the supply chain for chips was not so much a technology issue as a foundational economic and national-security issue. The supply chain for the components that go into drugs, when used by one country to constrain or have an impact on another country, becomes a tool of national power.

A software vulnerability is now not an IT issue.

Yep.

It can be a source of leverage, particularly if that vulnerability is in key parts of your power systems or your water systems.

Yeah. So I saw that shift happen, from tech being either an IT issue or just a tool to the central arena of both economic growth and national power. Quite frankly, from a deterrence perspective—which is the language we talk about between countries in terms of deterring an invasion of a country and maintaining an international system of stability—deterrence is no longer just the size of a military. I think, as we see in the Strait of Hormuz, or as we saw in the Red Sea, it's the pace of innovation in reacting to cheap, software-built technology that adversaries may be using, whether they're a country, a criminal, or a terrorist group.

Yep.

And those technologies today are not being built by governments. They're being built by the private sector. Whether that's AI systems, autonomous systems, or cyber-defense systems, they are being built in the private sector.

So, certainly from a U.S. perspective, we need to ensure that we have relationships between government and the private sector to leverage that for our own economic growth and as a force for good on the world stage. But also, with our allies, we build together, we fight together, and we ensure that together we can have the advantages of that—and, to the extent that our technology also represents our values—

It's important that the world run on American and Western tech. Yeah, actually, that's a great point on the values point, because, Ben, you've said before, no country wants to get colonized—

In cyberspace, right? Particularly because AI is now just getting built in the 6-block radius—

In San Francisco. Say a little bit more about that.

When AI becomes the interface for everything, every tool we have, every technology we use, and every kind of system we have—from your car to your refrigerator to your education system—is going to be built on these models.

The models are not objective. They have opinions. They have strong opinions on things like history, culture, ethics, and values. By the way, those are very different depending on where you are. The history of wars, the history of a country, its laws, and whether that country was a good actor or a bad actor are all very different depending on where you're getting the information from.

If you believe, as I do, that, for all our flaws, our fundamental values and ideas about everything from the way our economic system works to the rule of law have resulted in basically the best opportunities for humans in the world, then you want those values to be projected onto the world more than you want to be the recipient of some other values. Those values may have been good at a time or may have some merits, but overall, they don't give people a real opportunity.

Giving people a chance—that was the premise of the original announcement we did for our fund announcement earlier this year, and the importance of that, particularly extending that chance to America's allies as well. Raghu, I want to get you involved on this as well, because portfolio companies are increasingly going international way sooner. They're multi-country, multi-channel, and multi-product way sooner in the life cycle. How much of this is also accelerated by what's happening in AI, and just the ability to push out software updates immediately?

The reality is that products and technology travel faster than companies can. If you have a great product, everybody learns about it very quickly, instantaneously. You can access it through APIs wherever in the world you are, and you can build on top of those things.

So what happens is, for startup companies, usually in the olden days, you didn't care about going international until you had a few hundred million dollars in revenue, and so on. Now these companies have to go international sooner and sooner.

While the products are being discovered internationally sooner, to Ben's earlier point, it's also easier to adapt these products for local markets thanks to AI. As AI becomes the interface, it's much easier to localize and internationalize these products. But in order to capitalize on the opportunity at scale, things don't change, right? I mean, you still have to be in-region.

You still have to understand the market structure of the region that you want to operate in. You have to understand partnerships, and you have to build relationships. Lots of other countries have very relationship-based economies, not transaction-based economies. So it's this interesting blend where the technology moves really fast, but the way you do business hasn't changed that fast.

Yeah. Actually, a really great example of what Raghu was talking about: when we were in Mexico recently, we were meeting with the CEO of TelevisaUnivision, Alfonso de Angoitia, who is a great friend of the firm and one of our global partners. We were asking, “Gee, what are your struggles?” One of his challenges is that he really faces heavy competition from some of the global streamers, like Netflix or Disney, which have this U.S. technology that's very powerful.

However, he has this unbelievable Spanish-speaking content, but it was hard for him to get it to play on a global basis because it's in Spanish. It's a Mexican Spanish accent, which is different from an Argentinian Spanish or Colombian Spanish accent. He was like, “How should we solve this?” And we're like, “Well, we've got this great company, ElevenLabs. They can help you with it.” We put them together, and in a couple of months he was in France and Poland and all over the world with this incredible technology that had his actors basically acting exactly as they normally do, with their right kind of intonations, their right laugh—everything—but in different languages.

And he recently, based on that success, got a big deal with Netflix. And so here we have something where we can bring our partners, who have a great product that's culturally local but really, really interesting, and make it a global product by not competing with their allies but partnering with their allies. I think that's a lot of what the program is about.

And I want to tie what Ben just said back to the values point, because people often ask us, “Chinese open-source models—how do you compare and contrast?” So in that case, they didn't have to worry that when the content was dubbed into different languages, if the speaker said something sensitive about Tiananmen Square or sensitive regarding the Chinese government, using an American model, that content would be translated and changed along the way. They knew it would be true to what was said.

And that's what we mean when we talk about American values: not censoring, not injecting political will into how these models operate, in the way that one would contrast. Now, open source is something we deeply value because at the end of the day, there's all different kinds of models and all different kinds of price points, clearly. But that's really a good example of what we mean when we say American and Western values reflected in the models that we deploy around our infrastructure.

For sure, the telenovelas are spicy enough. I don't know how much Tiananmen Square needs to be introduced necessarily in the storyline, but you never know at the border crossings, et cetera.

Good point. They keep their audience well enough.

Exactly. Exactly. But this actually is a core part. When Piyush McCormack[?] wrote the article about us earlier this year, he called us a power broker, with power being part of an offering that we enable our founders to tap into for a lot of companies—

You know, starting a Mexico office, for example, is probably not 1, 2, 3—maybe even top 10—on their list, but if they have a deal that we're able to walk in—

—and effectively enable them a shot at winning way earlier in the lifecycle, they might prioritize that. So maybe if you want to riff on that, Raghu, Ben, particularly how that's changed for founders—how they think about international when they have the benefit of being able to go right to the buyers in a much more adept way than figuring out a market from scratch.

Yeah. A lot of markets are very top-heavy, right? So you take any mature business and you say, “How much of your revenue comes from countries like the U.K.?” A significant chunk is going to come from the very, very top companies in the U.K., right? So the reality is, whichever country you look at, it's 5 or 10 companies that matter. And, of course, the government is a big influencer and buyer.

So the example that you gave in Mexico is great, right? I mean, there are a few global companies in Mexico, like the one that Ben talked about, and so it's not worthwhile setting up an office and serving all of the countries, right? But if you could get into the top 5 companies in Mexico, that probably is worth a lot more than setting up a full operation in some other country. So it is the top-heavy nature and the relationship-heavy nature that helps our portfolio of companies, right?

And the thing—there are a couple of other aspects that are important. One is, you're right that Mexico probably isn't a top 10 thing that most companies go into first. And, by the way, that's a terrible shame, because Mexico is our neighbor. From a national security standpoint, how powerful would it be to have a Western Hemisphere supply-chain ecosystem that was really, really well connected? So I think it's kind of bad for the country that it is number 10.

Secondly, to open a country as a company is a very large expense. You have to set up an entity, you have to hire a team, and you have to localize the product. And that's going to be, depending on the country and depending on how committed you are to it, easily $5 million or $10 million just to get to your first deal. And so that is going to delay in your lifecycle when you can go international.

But what if, before you ever put up an entity or ever got started in that country, you had an opportunity—a big opportunity, a $5 million opportunity, a $10 million opportunity—that makes that investment much easier to do earlier? And so we can really enable both our companies and our allies by paving the way for that kind of thing.

A great example is, nobody's going to Saudi Arabia first. However, Saudi Arabia is booming, and there are great opportunities there, and they really need our help in developing the country. They're modernizing, which is incredibly important to the security of the world. Saudi Arabia is going from this kind of tribal, different kind of society, I would just say, to a modern, highly civilized player in the modern world and a full ally of the United States. That transition is really important to succeed.

So we're able to take a company like Flow, Adam Neumann's company, bring them to Saudi Arabia, introduce them to the right people in the government, introduce them to the right people in business, and introduce them to the right investors. It's the optimal place to live if you're an expat moving to Riyadh, helping develop the economy. And it's something that just isn't really possible without an effort like this.

For sure. Maybe let's talk about the geographies we're prioritizing, because there's a handful: Japan, Korea, the Middle East, Mexico, Canada. Are there themes as we think about different countries that make more sense versus others?

Yeah, this is timely. Raghu and I have been debating this. So I would just say, from our standpoint, it's very powerful when they're a strong ally of the U.S., we've got a relationship with the people in government, we've got a relationship with the people in business, and we've got a relationship with the investors. Those things are not separate; they all interact with each other. And if we have that, then we can really, really accelerate the opportunities for them and for us.

I think what I would just say to that point, building on Ben's point—and then I'll turn it over to Raghu—is what's changed so much, as we talked about a moment ago, is that on the global stage, whether it's in the context of countries pressuring other countries via supply chains, there are parts of the world where we have close allies who are now in increasingly tenuous positions.

So if we look at the Indo-Pacific region, China has vastly increased its military and vastly increased the number of nuclear capabilities that it has. Its shipbuilding capability is orders of magnitude greater than that of the United States, and our allies in the region have started responding. Japan has changed its really pacifist policy from the post–World War II period to say it must have a capable military.

Yeah. It wants to build the military to fight today's wars, not yesterday's wars. It doesn't necessarily have the military-industrial base. Now, that cuts both ways. As we know from our own kind of halting steps the United States military has been taking, the number of defense reform policies that have been written in the last 20 years, they all say the same thing, right? Fewer, smaller numbers of exquisite platforms; more, cheaper, larger numbers of platforms, in addition to very exquisite platforms for certain specific use cases.

So as Japan thinks about building a capable military in order to deter aggression in the region, it is looking to partner with new firms building autonomous ships, building AI for an integrated battlespace, and they want to build with us. They don't want to just buy, because they have their own capabilities they want in their own country and their own supply chains to protect.

So I think part of that is thinking about those regions of the world. I talked about one in the Indo-Pacific region that covers Australia as well. Certainly, the Middle East has been an area of American interest for a long time. Israel is an incredibly capable ally with a capable military and intelligence community. The UAE is as well.

Saudi Arabia is modernizing and thinking about AI infrastructure—how it builds an innovative, entrepreneurial innovation base—in order to compete globally. And certainly, when we think about Latin America, when we think about fentanyl coming into the United States, it comes in via Mexico. So public safety, national security, and how autonomous ships can help along a vast Mexican coastline are not only important to our ally and to the stability of Latin America; they're important to us as well.

So I think that's at least the national security lens we bring to some of our partnerships in American dynamism or in the AI infrastructure space.

Yeah. It's the reason Ben and I have been debating this: there are so many sides to it that you've got to bring it all together. Mm-hmm.

First, to Anne's point, it has to make a lot of sense from the point of view of America's interests, right? And that is abundantly evident in all these countries. These governments are very AI-forward, and they're putting in policies that are very modernization-centric, whether it's in defense or across the economy.

Thirdly, they are all top-heavy economies. I mean, when you look at the top companies, you are actually having global influence. For example, of all those Northeast Asian countries that you talked about, I think 15% to 20% of the Forbes Global 2000 are across Japan, Korea, and Taiwan.

And then lastly, the way business is done in these countries is very different from, let's say, the UK or some English-speaking country, right? So the value-add that we can provide to our startups and the value-add we can provide to these countries is so much higher than in other places.

Yeah. So when you put it all up and score it out, that's why we go into these countries.

Yeah. Some of these countries are also leapfrogging, where they didn't have to deal with the prior technology.

Exactly. In Saudi, you go through the visa entry, and it takes 30 seconds—30 seconds to go through the visa entry, which is unbelievable. Here, you get stuck behind, unfortunately, a very long line, increasingly for a number of different reasons.

It's a different kind of approach to how you think about incorporating technology in everyday life, but it's not evenly distributed. So what are ways we can accelerate that in different avenues? That is also the opportunity that a lot of these countries are seeing as a huge leg up.

It's very different, though, when we think about international prioritization compared to maybe other VC firms. The typical way that other VC firms have done international has always been investment-driven versus actually thinking about the portfolio companies and what they would like to do internationally.

So maybe speak a little bit more about that contrast and, quite frankly, why, for the first 10 years of the firm, it was very rare that we even did an international deal outside of not just the US, but even outside of Silicon Valley.

Well, I think we've always looked at investing through the lens of what we bring to the table. As a result, if you think about what we try to do for a company, we really try to help them network to the right talent, the right customers, and the right expertise. And that is very hard to do in multigeographies.

Even San Francisco and Menlo Park in the early days, we were like, "Let's just have a really good product for this, and then we'll expand from there." Whereas I think other VCs think of the product much more as the money.

And so, if the product is the money, then the international strategy is, "Okay, we'll find a team of investors in, say, the UK, and we'll help them raise money. Then we'll take a percentage of their winnings for helping get them their product." That makes sense in that model.

I think it doesn't make sense for us because our whole thing is, how do we help you build a better company? How do we accelerate your opportunity or growth and so forth? So we needed to get to a certain level of size and stature for that to make sense in more regions.

But now that we are, it starts to make a lot of sense. And I think, in our role in investing, we don't look at it as, "How do we snatch up the opportunities in Sweden or Japan or Saudi or the UAE?" We look at it as, "What is your ecosystem missing in terms of capital and expertise to go global, and is there an opportunity for us to plug in?"

We don't want to displace their A-round and seed investors necessarily. We want to say, "Okay, if you've got companies that are emerging as real global companies, then that's where we can really step in and help."

If we're already in the country, already doing a lot of business, and already understanding them and their entrepreneurs and so forth, then we're in a great position to do that. But we would do that directly. We're not raising money for somebody who's not part of our firm, our culture, or our mission and having them go run a play. That's just not us.

Also, so much of the time, when you think about these entrepreneurial ecosystems, what they're oftentimes missing is growth capital. They actually have a great talent base on the early-stage side, but they're kind of missing that component to actually go global, and they become a localized success versus actually thinking about global ambitions.

Maybe drawing from your experience at VMware, Raghu, about the evolution of company building, how has that particularly changed the ambition for local companies to become global over time?

That's dramatically increased, right?

Yeah.

Previously, like I said earlier, you would not even dream of going global until you were really well advanced. Now, because your product can be there, there's such a universal appeal for it. If you've got a great product, it has such universal appeal, whether it's a B2B or a B2C product, that it behooves you to go earlier.

What is missing is the supporting ecosystem, right? And finding a way to plug into a new country. That's where our new offering really focuses: helping them get a fast start in these regions.

For that, you need to be in the region, have the relationships, build the trust, and build the credibility. All of those things don't happen overnight, and that's where we can provide that.

And then, by spending time locally as well, we get to know the local ecosystem and see the great entrepreneurs from those regions. To the extent to which it makes sense, we fund those and then help them go international, leveraging our network as well.

So businesses in these countries are so much a collective ecosystem rather than an individual company going in and dominating something there.

For sure. For sure.

Maybe speak a little bit more about the prior tech cycles of adoption, and particularly with AI. Anne, this is a great question for you. There are so many nuances around security, and even the debate between open versus closed, that there are considerations of whether it's even viable to build without upsetting the apple cart on what is allowed with open source, given the security nuances.

Maybe speak a little bit more about how you've been thinking about that and advising companies on that delicate balance.

It's a hot topic right now, as you know, and it's a hot topic around the world because, fundamentally, as many technologies are, AI is dual-use.

The first step—let's use cyber as an example—the first step, whether you are an attacker or a defender, is finding vulnerabilities in systems. If you're a defender, you patch them. If you're an attacker, you exploit them.

Models make the step of finding vulnerabilities in code far easier.

Mm-hmm.

Quite frankly, they make the next steps, exploit or defend, easier as well. The economics of cyber defense were broken well before the latest AI models came out. I saw that in the White House, where my nightmare was the ransomware attacks against rural hospitals in America. There wasn't another hospital two blocks down.

So actual Americans' opportunity to access health care was disrupted by, in some cases, state-harbored criminal groups halfway around the world. Right? Infrastructure is very fragile, and the reality is that the models help both.

I personally believe, focused on cybersecurity for a moment, that they help far more in defense because defense is far harder. You have to be defending a broad expanse, and an attacker has to find one way in.

So, as a result, I actually think we are in a difficult space at this moment because there's such a broad attack surface and so many systems that are vulnerable. But I think we will see the opportunity for a much more secure space.

To your point, when I talk with governments or companies around the world, all are grappling with that question. I'll talk about the government side, given we know it can both help. Most countries have 2 missions, right? Defending the country, as well as maintaining its interests overseas.

Given both, how do we ensure that these powerful capabilities don't fall into our adversaries' hands or criminals' hands to be used for that purpose? I think, in the traditional cyber tools arena, we always grappled with this, right? When a capability is on a thumb drive, export controls are a joke. You can't actually control it in that way.

There is real power, as we talked about earlier, in the models that the world runs on being trusted models. When people are using an American model, they don't have to worry that there's an intentional backdoor. As we talked about earlier, they don't have to worry that a chatbot's answers will be framed by censoring to fit a particular political story of an authoritarian government.

So the question for now is: How do we build a degree of safety into models so that they can be used in environments that need that safety? Because the main models are American, we're in a particular position to think through and build that out. I think this is not done by government, and it's not done by the private sector. What it really needs is people who understand the models best from the private sector at the table with folks on the government side who are thinking, "How do we both promote innovation and control for these risks?"—and building out the right set of government policies to enable both the use of the models as well as the understanding of how to protect against the risks that come with them.

Yeah. As our partner Mark often says, technology is the dog that caught the bus. We've never seen this before: countries feverishly demanding the latest versions of the model because they recognize how competitive it is for their economy, their people, and their interests.

This is unlocking all sorts of new paradigms as we think about globalization in a way that feels categorically different from just selling products internationally, right? This is hopefully an opportunity for many of our founders as well.

No doubt. Although it is complicated. For example, from a government standpoint, you'd like to say, "Okay, don't release your model with cyberhacking capabilities." But if you do, and you want to find the vulnerabilities in your own code and patch them, you can't do that without also enabling cyberattacks, because as soon as it can find the vulnerabilities, guess what?

And so you're in a very—

I would say it's an interesting world on that.

Yeah. And I think, to that point, we'll focus on the cyber aspect because that's gotten so much attention with Anthropic's Mythos model, etc.. I think what AI models essentially do, to Ben's point, is allow you to be the first to hack your systems.

We often use the expression "digital doorknobs." It now allows attackers to jiggle every doorknob continuously and at scale. So your hope for your network is to use those models first to find where you're vulnerable and, frankly, then be able to fix it. That wasn't a tractable problem before.

I think now the models help you, particularly for code bases that were public, particularly if you're relying on third-party code or open-source code. Those are some of the biggest areas of risk, or where there are dependencies on code that nobody was really maintaining.

We're now seeing companies come together and take responsibility for that, because the cost of not addressing insecure code has gone up and the cost of addressing it has gone down.

Yeah.

Which is a good shift from where we were before.

We may finally build secure systems.

Exactly.

Please God.

And more importantly, the process for building new code can incorporate these as well.

It's kind of very tricky from an industry standpoint, because originally cyberhacking was vandalism, right? "Okay, I'm going to mess up your PC." Then, with the internet, it became theft, and now it's war.

That evolution, I think, snuck up on people.

We've seen the great thing when it went from vandalism to theft: The vandalism company, Microsoft, is the one that created all the theft opportunities by putting Windows on the internet when it wasn't ready to be on the internet, because it wasn't built for that kind of environment. With ActiveX and all these kinds of technologies, people could come in from the internet and get right to your machine.

I think we're in a period like that, where we've got all this technology that was built in the pre-AI era that's not nearly secure enough for the AI world. We're going to have to make that transition.

Yeah. There's a 20-year legacy of code that has to be tackled.

Yeah.

Nope.

Many opportunities.

Yes, many opportunities. Yeah.

Many, many.

Awesome. Well, maybe in closing, Ben, I know the number-one question we often get when we're visiting different countries is, "How do you recreate Silicon Valley locally?" They want the ingredients to it. There are a few components to the success that enabled Silicon Valley, but there's also the proposition: What if the internet is actually Silicon Valley, or Silicon Valley is just distributed?

Yeah. Yeah. I don't think that's true, by the way. They think that's happy talk. One thing you learn when you go to different countries is that they're very different. The laws are different, the policies are different, and the governance is different. That's why America has continued to be special. We've had the internet now for quite a long time, and that's absolutely not the case.

I wish we could bring some of our own policymakers on those trips so they could hear people say, "We want Silicon Valley," before they keep trying to wreck us. Not all of our policymakers—many are outstanding—but some really don't like the fact that we have a technology industry.

As we've traveled around, there's a combination of ingredients, and they are a little bit the combination that makes Silicon Valley special and America itself special. The first is that you do need the talent. Talent is most easily seen in whether there's a great technical university that graduates people who know how to build things. That's an essential thing. Many countries have that and still don't have a Silicon Valley.

What are they missing? They're missing the other 2 components. The second is whether you have a set of laws and policies that facilitate entrepreneurship and are essentially good for business. Can you create a company easily? Can you hire people easily? Can you fire people easily? Is there a tax incentive structure that's not onerous?

Can you own—and this is a big one lately, which California is kind of moving against—an illiquid asset that has value, or will it get seized? Norway introduced an unrealized capital-gains tax, and that effectively ended the tech ecosystem there. Those kinds of policies are super important. If they're not right, nobody is going to take the extreme personal and financial risk of starting a company if it's not an environment that's conducive to that.

The third one, which is a combination of everything in a way, is whether the culture is such that young people—the most capable young people, willing to make the biggest contribution—get social-status rewards. Is it cool to either start or join a new company? Very few countries have that. I would say it's really, really good in the U.S. It's clearly good in China, by the way, although they've done a few things to mess it up.

But in most countries, success is frowned upon. You hear a lot of this in the political rhetoric: "Oh, the only way you could make a billion dollars is if you're a thief," which is completely absurd. You just look at the people—almost none of them are thieves. You can't steal a billion dollars; that's a real truth.

It's a great sour-grapes idea, right? It's like, "How did he get all that money? Oh, he's a thief. Okay, I'm not a thief, so I feel much better about myself now." But that's a lot of it, and that's the thing that's so hard to replicate, so hard to build, and so easy to destroy.

It's something that I think we have to be aware of: how special and unique that is in the world. We can try to help other people do it, but we also have to try equally hard not to wreck it ourselves.

Yeah, for sure. It's World Cup season, and there was a great tweet where someone was like, "I think America thinks they could potentially win the World Cup." Maybe they can, maybe they can't, but the fact that America thinks they can is a feature, not a bug.

Yes.

There’s a reason ambition is fantastic.

There’s a reason we all read to our kids. What is it? “The Little Engine That Could.” I think I can. I think I can. I think I can. That’s in America. You can believe you can.

That’s right. That’s right. Amazing. That’s a good note to end on. Thank you all.